Use of Artificial Viscosity with PyFR

[Antonio Garcia-Uceda]

Dear all,

I'm trying to run the test case: Inviscid transonic flow around NACA0012, from the previous 3rd High-Order Workshop (you can find a description below):

https://www.grc.nasa.gov/hiocfd/wp-content/uploads/sites/22/case_c1.2.pdf

I've run the case on a coarse mesh of 2240 cells, order p3, and the parameters for the artificial viscosity method specified in the "User Guide" section of your website: 

[solver-artificial-viscosity]
max-artvisc = 0.01
s0 = 0.01
kappa = 5.0

I obtain indistinguishable results from those without the artificial viscosity, so I guess I need to augment the max-artvisc value in order to get a higher smoothing of the shock.

Could you please let me know, first, if you have run this test case or another similar, and second, which values for the parameters above to use?

Also, for a more general test case, Which procedure do you use to estimate the correct value of the parameters above? I guess it's at the end a process of trial and error, but Would you know how to do a first quick estimation?

Thanks a lot for your help.

Best regards,

Antonio 

[Vincent, Peter E]

Hi Antonio,

What system are you solving? And can you confirm that you have shock-capturing = artificial-viscosity

Cheers

Peter

Dr Peter Vincent MSci ARCS DIC PhD
Reader in Aeronautics and EPSRC Fellow
Department of Aeronautics
Imperial College London
South Kensington
London
SW7 2AZ
UK

web: www.imperial.ac.uk/aeronautics/research/vincentlab
twitter: @Vincent_Lab

--
You received this message because you are subscribed to the Google Groups "PyFR Mailing List" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pyfrmailingli...@googlegroups.com.
To post to this group, send email to pyfrmai...@googlegroups.com.
Visit this group at https://groups.google.com/group/pyfrmailinglist.
For more options, visit https://groups.google.com/d/optout.

[Antonio Garcia-Uceda]

Dear Peter,

I'm solving the Euler equations (or Navier-Stokes with mu=0.0 since the artificial viscosity model only works for the Navier-Stokes solver, as far as I know).

Antonio

[Antonio Garcia-Uceda]

Dear Peter,

This is a reminder for the email above from previous week.

I'd be very grateful if you or your team could provide some guidance on the use of the "artificial-viscosity" tool in PyFR, in particular on how to tune the parameters of the scheme.

Many thanks in advance.

Best regards,

Antonio